Potentially inflammatory rhetoric should be addressed with dialogue, not bans

Recent antisemitic attacks have rightly led to concern among the US Jewish population. Over the weekend a man assaulted worshippers at a New York synagogue in an antisemitic attack. A man stabbed a Jewish man outside another New York synagogue that was displaying the Israeli flag. In separate incident, a man threw a brick through a Florida Chabad House. We’ve also seen incidents here in Vermont, with a synagogue vandalized with “Free Palestine” last year.

Jewish groups are also rightly debating who is responsible for such attacks. Obviously, the attacks are ultimately responsible. But there is a fear that a broader climate of antisemitism, connected to criticism of Israel, is leading to these attacks. After a separate antisemitic vandalism incident in Vermont, the victim said that attacks on Israel by Vermont groups has led some to expand their targeting to all Jewish people.

And much of this has focused on New York City Mayor Mamdani, who has been a strident critic of Israel. Some Jewish groups have argued that his rhetoric has emboldened individuals to carry out recent antisemitic attacks. The NYPD has said this will heighten security risks when Netanyahu visits New York for the UN General Assembly meeting this fall. Specifically, Mamdani had pledged during his campaign to enforce an ICC warrant for Netanyahu’s arrest, and recently called on the US government to join the ICC and arrest Netanyahu.

This debate calls to mind the idea of “stochastic terrorism,” which has been popular among progressives in recent years. As Vox put it, this is the idea that rhetoric short of direct incitements to violence can still lead to random acts of terrorism when they “[entertain] it as a legitimate tactic” or “[demonize] a political enemy.”

This has usually been associated with right-wing actors, but could Mamdani be credibly accused of stochastic terrorism? Based on the definition of stochastic terrorism, yes. But we need to think deeply about how best to address it.

What is stochastic terrorism?

It’s worth expanding on this term, because, despite being a popular buzzword, it’s applicability in practice is a little unclear.

In the Vox article, a communications professor says stochastic terrorism occurs when individuals use mass communication channels to imply the need for violence against a group or individual without explicitly directing it. This allows for plausible deniability. A Mother Jonesarticle defined it a little more broadly, saying it’s “pinning a bull’s-eye on the back of an opponent in a volatile situation—perhaps suggesting the world would be safer without this supposed threat—knowing this could lead to violence against that target.” The author gives an example of Elon Musk accusing George Soros of trying to destroy Western civilization.

There is some conceptual slippage in these popular discussions. At times it is general demonization of an individual or group, while at other times it is specific indirect calls for violence. Academic discussions tend to focus on the former: as one study put it, “the incitement of a violent act through the public demonization of a group or individual.” I think this is the best approach. Incitement to violence is already a thing, and debates about what counts as direct or indirect incitement should fall under that category.

There don’t seem to be many empirical studies of stochastic terrorism, but academics seem to accept it as plausible. One study discussed mass psychological processes through which it could occur. A critical terrorism studies article argued it is related to authoritarian and racist impulses in liberal democracies. A legal study argued that US groups engaging in stochastic terrorism should lose their tax exempt status. Others discussed stochastic terrorism as arising from “demagoguery.”

Is Mamdani a stochastic terrorist?

Based on this definition, I’d argue Mamdani’s rhetoric could be called stochastic terrorism.

Yes, he did not directly target Jewish people with his speech, but he is smart enough to know there are extremists who conflate Jewish people with Israel. So criticism of Israel would lead to negative views of Jewish people. And antisemitic attacks have increased in the wake of his rhetoric.

While Mamdani has not engaged in hate speech, he’s arguably engaged in demagoguery. According to Merriam Webster, this is behavior resembling a demagogue, “a political leader who appeals to popular prejudices and who makes false claims and promises in order to gain power.” Mamdani, again, has to be smart enough to know he had no power to arrest Netanyahu under an international criminal court warrant the US is not party to. And he had to have known calling on Trump to join the ICC would accomplish nothing. He did this to satisfy his followers and demonstrate his credentials on this issue.

Like many neologisms, it’s not clear how novel “stochastic terrorism” is.

And there is some evidence of a direct connection between his rhetoric and antisemitic violence. The Florida incident involved a man who had previously shared Mamdani’s call to arrest Netanyahu and who said social media had led him to commit the attack. This seems like a pretty clear connection.

Imagine a similar scenario on the right. A right-wing politician is elected partly on a platform of arresting a progressive foreign leader, and later calls on the federal government to take action against that leader when they are in the United States. An individual then attacks a site that, at least in the individual’s mind, is associated with that leader. I’m sure progressives would denounce the right-wing politician as a stochastic terrorist.

There are obvious counterarguments. Mamdani has been clear that he sees a distinction between criticism of Israel and antisemitism, suggesting it was not his intent to cause harm to Jewish people. One could argue there is a difference in kind between Mamdani’s criticism of Israel and Netanyahu and the sort of rhetoric people point to as stochastic terrorism on the right.

The issues with stochastic terrorism

Issues remain with accepting stochastic terrorism as a valid term, however. It is a complicated concept, as it implies both causality and culpability: an individual demonizing a group or individual causes acts of violence against them, and is implicated in it.

The first issue, as I’ve already implied, is empirical. I haven’t seen convincing studies demonstrating stochastic terrorism exists (send any I’ve missed). I guess it’s possible to conduct such a study: you could look at acts of violence against a target group after a “stochastic terrorist statement.” It’d be tough to demonstrate causality, and culpability would be hard to test; you’d have to prove the speaker’s deep-seated motivations, which is never easy.

The other is conceptual. Much of the discussion of stochastic terrorism, even among academics, is value-laden. The critical study I cited says this arises from the “mainstreaming of racism.” The legal study said that the “law is not yet enlightened,” and explicitly conflates hate groups and stochastic terrorism. The use of “demagoguery” to define it is tricky, as this is a vague term one only applies to their political opponents. Thus, many scholars of stochastic terrorism assume it is confined to right-wing hate groups, and then using this as evidence for their arguments. But taking the concept on its own, it’s not clear why it would have to be right-wing. Anyone can demonize a political opponent, leading to violence.

Instead of banning speech as “stochastic terrorism,” we need to start with a responsibility on the part of both the speaker and the concerned audience.

Moreover, like many neologisms, it’s not clear how novel “stochastic terrorism” is. Of course rhetoric from influential people can inspire some violence among a small sub-set of society. Many people read Marx, but only a few became Marxist guerillas. Many people read Sayyid Qutb, but only a few became Islamist militants. It’s very difficult to predict who will become a terrorist, and ideological fervor tends to be an exceptionally weak predictor. So the idea that rhetoric could lead some individuals—possibly already pre-disposed to extremist violence—to commit terrorism isn’t anything new. The difference, as I’ve suggested, is assigning this to only one political orientation and saying this leaves the initial speaker complicit in the violence.

Finally there’s the question of what to do about it. Most of the arguments say we need to condemn it. Some call for legal actions. And the UK attempt to ban anti-Israel speech on public safety grounds seems like a logical extension of concerns about stochastic terrorism. As I’ve discussed before, I’m wary of that approach. This creates a bit of an inverse of the heckler’s veto. The heckler’s veto is a speech restriction due to the possibility of a violent reaction to the speaker. In this case, it would be a speech restriction due to the possibility of what a supporter might do. In some cases it’s easy to determine, but much of the time it would be hard to predict.

It’s for these reasons—empirical, conceptual, and practical—that I’ve always been skeptical of the term. I don’t like hateful speech—hot take—but we should counter the speech’s content, not try to shut it down with a new category of banned rhetoric. We should study the connections between rhetoric and terrorism, not assume the former is responsible for the latter. And any concept that’s meant to be universal but smuggles in one-sided political arguments is flawed.

Moreover, the counterarguments to my claim that Mamdani’s rhetoric counts as stochastic terrorism just highlight the issues I noted with the term. If it is reliant on value judgments about what criticism is valid, or which targets are valid, it’s not an empirically sound concept. If it relies on understanding the inner motivations of a speaker, it will always depend on subjective perceptions of that speaker.

So what to do?

This creates a dilemma. I’m arguing that, based on a consistent definition of stochastic terrorism, Mamdani’s rhetoric would count. But I also think that this concept is flawed, and not particularly useful in understanding political violence. So it’d be concern trolling for me to argue that Mamdani is implicated in recent antisemitic attacks due to stochastic terrorism.

I’m not going to go so far as to say we should dismiss any concerns about Mamdani’s rhetoric, however. I tend to be on the side of free speech, but I also think people have responsibility for their words. If I say something and someone tells me it offended them or reminded them of dangerous rhetoric they’d heard in the past, I should apologize and think about what I say in the future. Likewise, if Jewish groups raise concerns about Mamdani’s rhetoric, he should reflect on its impact. To his credit, he has condemned antisemitic violence, but has been criticized for a “slow, muted” response over an intimidatory pro-Hamas rally outside a synagogue. He has a right to criticize Israel, but should also take think about how to do so in a way that calms constituent concerns.

So what should be done about this? I’d call for a dialogue-based approach, drawing on another area I follow, international religious freedom promotion. As I’ve written, the IRF community has focused on condemnation of abuses, which has led them to a fruitless alliance with the Trump Administration. An alternate approach is what Chris Seiple called the “builders” approach, in which those concerned about rights abuses work with those committing them to come to a mutual agreement. It’s not easy or satisfying, but may be most effective. Seiple’s paradigm is focused on government engagement, but it reminded me of Loretta Ross’ “calling in” paradigm for inter-personal relations. It also reminded me of Daryl Davis’ work engaging with an converting white supremacists away from their hate. And I’d argue we could even see it in Mamdani’s meeting with Trump, despite the nasty things Trump had said about him (which could have been termed stochastic terrorism); Mamdani chose to engage with someone who had used dehumanizing language about him.

Instead of banning speech as “stochastic terrorism,” we need to start with a responsibility on the part of both the speaker and the concerned audience. The speaker needs to realize that they are causing concerns with their rhetoric, and should be receptive to outreach by the audience. And the audience needs to be willing to engage with the speaker to find common ground.

When it comes to extreme cases like the demonization of Haitians during the 2024 US election, it’s hard to imagine those spreading the hate backing down in the face of engagement (although Davis’ work suggests there is a chance). But in the case of Mamdani’s rhetoric, there is an opportunity. I don’t think he wants to bring harm to Jewish people, but I do think he de-emphasizes antisemitism when it comes to controversies involving Israel. He probably won’t stop criticizing Israel, but Jewish groups (rightly) won’t stop expressing concern about rising antisemitism. There should be a way to meet and figure out a path forward. The recent meeting between New York rabbis and Mamdani is a good example of this; reports suggest Mamdani seemed open to the rabbis’ arguments, and the rabbis suggested concrete ways to move the relationship forward.

I will close by noting it’s easy for me to say this. I’m a white Christian guy who is not personally affected by either antisemitism or Israel’s violence against Palestinians. If I was just sitting on the outside criticizing those in the midst of these struggles, that would not be ok. But I base my views on the insights of people who come from communities dealing with these issues (I have a working paper based on this I’m trying to finish soon). So I see it as an authentic desire for those within these struggles that I, as an outsider, hope to amplify.